The abstract for a new arXiv paper argues that the idea presented in a published paper is seldom the one a research project started with. Instead, the authors describe a closed-loop process in which researchers run experiments, derive insights, identify challenges, explore the literature, and revise their thinking along the way.
To capture this dynamic, the paper proposes a graph-based view of research ideation. Rather than treating a paper's final contribution as a static endpoint, the model maps the evolving relationships between questions, findings, and conceptual shifts. This framing makes it possible to study how authentic research trajectories unfold.
The work is primarily a conceptual and methodological contribution, offering a way to represent the messy, iterative nature of discovery. It may be useful for researchers studying the science of science, as well as for tools that aim to support or automate aspects of the research process.